JPH0312695B2 - - Google Patents

Abstract

A fiber optic sensor comprises a length of optical fiber (10), forming a loop (40), and a fiber optic directional coupler (20) for optically closing the loop (14). The loop (14) and coupler (20) form a resonant cavity for light circulating therethrough. A PZT cylinder (118), about which the fiber loop (14) is wrapped, is utilized to control the total round trip phase delay of the circulating light, and thus, control the intensity of the optical output signal W_o. The phase delay is adjusted to a point Z where the optical output signal W_e is at maximum sensitivity to changes in phase. When the fiber loop (14) is exposed to, e.g., acoustic waves, the loop length changes correspondingly, thereby causing the phase delay, and thus, the optical output signal W_o to vary. By detecting variations in output signal intensity, the frequency and intensity of the acoustic waves may be determined. The sensor also includes a feedback system for stabilizing the fiber loop (14) against low frequency thermal drift.

Description

[Detailed description of the invention]

BACKGROUND OF THE INVENTION This invention relates to fiber optical sensors, and relates to stabilized fiber, resonant ring interferometric sensors. Optical fibers are sensitive to a large number of environmental effects, such as acoustic waves and temperature fluctuations.
When environmental quantities act on an optical fiber, the amplitude, phase, and
Changes occur in the state of polarization, etc. This has led to increased interest in recent years in using fibers as sensor elements. Commonly used sensors include Mach-Zehnder and Sagnac interferometers. In these interferometers, the phase difference between the interfering light waves changes in response to changes in the sensed quantity, so it is necessary to detect the intensity of the light produced by such interfering waves. can determine the magnitude of the sensed quantity. Also, the detected intensity varies as a cosine function of the phase difference between the light waves. Typically, this interferometer is biased to operate at the point of maximum slope above such a cosine curve, and the sensed quantity is detected by measuring the fluctuation in intensity from this operating point. be done. However, the maximum slope of the cosine function is approximately 1/2 radian ^-1 when the peak-to-peak optical power is normalized between 0 and 1. This limits the sensitivity of such devices. Furthermore, since the cosine function has a strongly nonlinear nature in the region toward its minimum and maximum values, generally only a portion of this curve can be used as a linear measurement. SUMMARY OF THE INVENTION The present invention comprises a fiber optical sensor comprising a single seamless strand of optical fiber having first and second ends and a loop portion therebetween. Contains. This loop is closed using a fiber optic coupler. A light source is used to provide a light wave at the first end, which propagates through the fiber to a second end.
reaches the end of the loop and provides an optical output signal, the intensity of which is dependent on the length of the loop. The circular phase delay of the light rotating through this loop is controlled by a phase retardation device to provide a resonant cavity for the rotating light. The fiber loop length is sensitive to first and second environmental effects, the first of which varies within a first frequency range and the second of which varies within a second frequency range. Varies within a frequency range. The phase delay device is driven by a stabilizing device to compensate for fiber loop length changes caused by the first environmental effect and to stabilize the output signal with respect to the first environmental effect. However, during such stabilization, the output signal remains sensitive to secondary environmental effects, such that the detector reduces the intensity in the output signal. is detected by measuring the variation in Preferably, the phase delay device provides a full round phase delay for the rotating light such that the sensitivity of the output signal to changes in loop length is at a maximum. The phase retardation device typically includes a mechanism for stretching the fiber, such as a PZT cylinder around which a portion of the fiber loop is wound. The stabilization device preferably includes an electronic circuit that drives the phase delay device with an electrical signal from the detector. In such a case, the electronic circuit filters the signal to pass frequency components within the first frequency range and effectively blocks frequency components within the second frequency range. In one preferred embodiment, the electronic circuit includes an integrator that selectively provides gain to signal components within a first frequency range. The device described above can be used in a method of sensing acoustic waves. This method involves forming a loop in an optical fiber and providing input light,
It involves propagating this light through a loop to provide an optical output signal. A fiber optic coupler is used to optically close the loop. The length of the loop, which affects the strength of the output signal, is chosen to provide a resonant cavity. Thermally induced loop length changes can be compensated for by stabilizing the output signal strength relative to the thermally induced changes. This output signal is detected to sense the presence of acoustic waves. Preferably, a point of maximum sensitivity of the intensity of the optical output signal to a change in the loop is determined, and the length of the loop is chosen such that the intensity of the output signal is at this point of maximum sensitivity. In a preferred embodiment, compensation for thermally induced loop length changes is performed by detecting components of the output signal having a frequency lower than the frequency of the sensed acoustic wave, and The loop length is subject to change in response to detected frequency components. In particular, the sensor of the present invention alleviates the problems associated with prior art techniques utilizing high finesse all-fiber optical ring resonators and has a very sharply spiked and very linear output power curve. Preferably, the slope of such a power curve is on the order of 14 radian ^-1 with a resonator finesse of about 80.
This particular type of resonator thus provides a highly accurate fiber optical sensor for measuring environmental quantities that affect the light propagating therethrough. In such cases, the length of the loop changes as a function of the magnitude or frequency of the acoustic wave entering the loop.
Such a change in loop length then causes the output power of the resonator to change accordingly. By detecting this change in output power, the magnitude and frequency of the acoustic wave can be directly indicated. Changes in the length of this loop can also be caused by other environmental or ambient influences, such as temperature fluctuations, which cause the output power of the resonator to drift at relatively low frequencies. In the preferred embodiment of the invention, a feedback system is used to compensate for such low frequency drift. Preferably, the observed sensitivity of periodic phase shifts induced by changes in loop length in the fiber ring is 1.0×10 ^-6 rad/√ over the frequency range of 100Hz to 10KHz. Hz or higher, and is 1.0×10 ^-7 rad/√Hz at 10KHz. In the region above 100 Hz, the sensitivity of this preferred embodiment is at least twice that of the corresponding Matsuzher-Zehnder interferometer. Testing of the preferred embodiment shows that this sensitivity is not limited by shot noise, but rather by the spectral bandwidth of the single frequency laser source. Therefore, improvements in the spectral purity of single frequency lasers should increase the sensitivity described above. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The fiber optic laser of the present invention utilizes a fiber optical resonator. As shown in FIG. 1, the resonator comprises a continuous, uninterrupted strand 10 of single mode optical fiber;
This strand has an input end 12 and a loop part 1
4, and an output end 16. At each end of the loop portion 14, a fiber 12 is optically coupled thereto, with ports 1 and 2 on one side and ports 3 and 4 on the other side.
This is done by a port directional coupler 20. Tracing from one end of the fiber 12 to the other,
Fiber 10 first passes through ports 1 and 3 and then through ports 2 and 4. By this,
Loop 14 extends from ports 3 and 2, while
Input portion 12 would extend from port 1 and output portion 16 would extend from port 4. A laser light source 2 is used to enter the input fiber section 12 and rotate the light through the loop section 14.
2 is used. This light source 22 is in Tropel mode.
A single frequency, long coherence length, continuous wave laser 24, such as a 1200 He--Ne gas laser, is included. This laser has a wavelength of 632.8nm and a
It generates single frequency light with a coherence length well over a kilometer. The light from the laser 24 is focused on the fiber section 1.
2, a lens 26 can be used. In addition to this, the laser 24 and lens 2
6, a polarizer 30 and a quarter wavelength plate 3
2 may be inserted to prevent light reflected from lens 26 and fiber section 12 from entering laser 24 and interfering with its operation. In the resonator shown in the example, the fiber 10 comprises an ITT single mode fiber with a core diameter of 4
micron, effective core area of 1.6×10 ⁻⁷ cm ² , effective reflectance of 1.46, and absorption of 8.3 db/Km. Loop 14 includes a polarization controller 40 to compensate for fiber birefringence within loop 14, thereby causing port 2 of coupler 20 to
The rotating light at will have substantially the same polarization as the light from the laser source 22 at port 1. Polarization controller 40 and combiner 20 are fiber optics, distinct from bulk optical components. Coupler 20 A preferred fiber optical directional coupler that can be used as a coupler in a resonator is the 1980
Electronics Letters Volume 16, No. 7, published March 29,
It is described on pages 260 to 271. As shown in Figure 2, this coupler:
It includes two exemplary strands 50A and 50B of single mode fiber optic material, which are mounted vertically in grooves 52A and 52B, respectively. These grooves 52A and 52B are
Each is formed on optically flat opposing surfaces of rectangular bases or blocks 53A and 53B. Connect block 53A with strand 50A installed in groove 52A to one side 51 of the coupler.
The block 53B with the strand 50B installed in the groove 52B is called A and is connected to one side 51B of the coupler.
I will call it. The arched grooves 52A and 52B have a very large radius of curvature compared to the diameter of the fiber 50 and a width slightly larger than the diameter of the fiber so that when the fiber 50 is installed therein, , is adapted to be fixed in a path defined by the bottom wall of the groove 52. The depth of grooves 52A and 52B varies from a minimum at the center of blocks 53A and 53B to a maximum at the edges of blocks 53A and 53B, respectively. This is due to the fact that when the fiber optic strands 50A and 50B are installed in the grooves 52A and 52B, respectively, they gather loosely toward the center and block the blocks 53A and 53.
This gradual widening towards the edge of B prevents any sharp bends or sudden changes in the direction of the fiber 50 that could cause losses through mode perturbations. It has Although the groove 52 is rectangular in cross-section, it may alternatively have any other suitable cross-sectional profile convenient for the fiber 50, such as a U-shaped cross-section or a V-shaped cross-section. Good things will be understood. In the embodiment shown here, in the central part of the block 53 the depth of the groove 52 in which the strand 50 is attached is smaller than the diameter of the strand 50, while at the edge of the block 53 the depth of the groove 52 is smaller than the diameter of the strand 50. The diameter is preferably at least as large as the diameter of the strand 50.
The fiber optic material has been removed from strands 50A and 50B, for example by wrapping, so that each has an elliptical flat surface that is similar to block 5.
It will exist on the same plane as the opposing surfaces of 3A and 53B. These elliptical surfaces from which the fiber optical material has been removed will be referred to herein as the fiber "facing surfaces."
The amount of fiber optic material removed therefore increases slowly from O in the direction of the edges of block 53 and reaches a maximum in the direction of the central portion of block 53. Tapering the fiber optic material allows the fiber to gradually narrow and widen to prevent excessive back reflections and loss of optical energy. In the embodiment shown, the coupler sides 51A and 51B are identical and are placed in such a way that the surfaces of blocks 53A and 53B are opposite each other and the opposing surfaces of strands 50A and 50B are in facing relationship. They are assembled by arranging them so that they have . An index matching material (not shown), such as index matching oil, is placed between opposing surfaces of block 53. This material has a refractive index approximately equal to that of the cladding, and also serves to prevent optically flat surfaces from becoming permanently locked together.
This oil is transferred to block 53 by capillary action.
be placed between. Interaction regions 54 are formed at the junctions of strands 50 where light is transmitted between the strands by evanescent field coupling.
To ensure proper evanescent field bonding, care must be taken to control the amount of material removed from the fiber 50 such that the distance between the core portions of the strands 50 is within a predetermined "critical zone." It has been found that it is necessary to control the This evanescent field extends further into the cladding and decreases rapidly with increasing distance outside their respective cores. Therefore, a sufficient amount of material must be removed to ensure that each core is substantially within the evanescent field of the other. If too little is removed,
The proximity between the cores is not sufficient and the evanescent field coupling does not provide the desired interaction between the guided modes, resulting in poor coupling. Conversely, if too much is removed, the propagation properties of the fiber are altered and optical energy is lost due to mode perturbations. However, if the distance between the cores of the strands 50 is within the critical zone, each strand can receive a significant portion of the evanescent field energy from the other strands without energy loss becoming an issue. A strong bond can be achieved. This critical zone includes regions of sufficient overlap such that bonding between the fibers is achieved, ie each core lies within the evanescent field of the other. However, as previously shown, if the cores are too close together, modal perturbations will occur. For example, for weakly guided modes such as the HE ₁₁ mode in single mode fibers, such mode perturbations occur when enough material is removed from the fibers 50 to expose their cores. It is considered a thing. The critical zone is thus defined as the region where the evanescent fields overlap with sufficient strength to produce coupling such that no substantial modal perturbation-induced power loss occurs. The extent of the critical zone for a particular coupler depends on a number of interrelated factors, such as the parameters of the fiber itself and the geometry of the coupler. Furthermore, for a single mode fiber with a step index cross section, this critical zone can be very narrow. In single-mode fiber couplers of the type shown here, the center-to-center distance between the strands 50 in the center of the coupler is typically greater than several times the core diameter (e.g., 2-3 times). Something small is required. Preferably strands 50A and 50B
(1) are identical to each other; (2) have the same radius of curvature in the interaction region 54; and (3) have an equal amount of fiber optic material removed therefrom to form their respective opposing surfaces. ing. Thus, the fiber 50 passes through the interaction region 54.
Their opposing surfaces are symmetrical,
Therefore, their opposing surfaces are coextensive when superimposed. This allows the two fibers 50A and 50B to connect to the interaction region 5.
4 will also have the same propagation properties, thereby preventing coupling absorption that would occur if the propagation properties were different. Block or base 53 may be made of any suitable rigid material. In the presently preferred embodiment, base 53
is about 1 inch long, about 1 inch wide and about
It typically comprises a rectangular block of fused silica glass having a thickness of 0.4 inches. In this embodiment, fiber optic strand 50 is secured within slot 52 by a suitable adhesive such as an epoxy adhesive. Fused silica glass block 5
One of the advantages of using block 53 and fiber 50 is that it has a coefficient of thermal expansion similar to that of glass fiber, which means that block 53 and fiber 50 are This is particularly important when undergoing heat treatment.
Another suitable material for block 53 is silicon, which also has excellent thermal properties for this application. The combiner of FIG. 2 includes four ports labeled A, B, C and D, which correspond to ports 1, 2, 3 and 4, respectively. Viewing FIG. 2 as a whole, ports A and B, corresponding to strands 50A and 50B, respectively, are on the left side of the coupler, while ports C, corresponding to strands 50A and 50B, respectively, are on the left side of the coupler.
and D are on the right side of this coupler. For convenience of discussion below, it is assumed that input light is provided to port A. This light passes through this coupler to port C
and/or output from port D, which output is dependent on the amount of power coupled between the strands 50. In this regard, the “coupling constant” is
Defined as the ratio of combined power to total output par. In the example described above, this coupling constant would be equal to the ratio of the output of port D to the sum of the power outputs at ports C and D. This ratio is also referred to as "coupling efficiency" and is usually expressed as a percentage when this term is used. Therefore, when the term "binding constant" is used herein, the corresponding binding efficiency must be considered to be equal to this binding constant multiplied by 100. For example, a binding constant of 0.5 is equivalent to a binding efficiency of 50%. The coupler allows the coupling constant to be "adjusted" by offsetting the opposing surfaces of block 53 to any desired value between 0 and 1.0. Such adjustments can be made by sliding the blocks 53 horizontally relative to each other. This coupler is highly directional; almost all of the power applied to one side of the coupler is transferred to the other side of the coupler. That is, almost all of the light applied to input port A is transmitted to ports C and D, and no reverse coupling to port B occurs. Similarly, almost all of the light applied to port B is transmitted to ports C and D. Furthermore, this directionality is symmetrical, so that almost all of the light applied to either port C or input port D is transmitted to port A.
and communicated to B. Furthermore, this coupler makes little distinction between polarization states, so
It has the property of preserving the polarization state of light. Therefore, for example, if a vertically polarized light beam enters port A, the vertical polarization of the light traveling from port A to port D due to cross-coupling will be the same as the light traveling straight from port A to port C. is maintained. Light cross-coupled from one fiber to another experiences a phase shift of +π/2, but
Non-cross-coupled light does not shift out of phase while propagating through this coupler. Therefore, for example, when light enters port A, the phase of the cross-coupled light at port D advances by π/2, while the phase of light traveling straight to port C remains unchanged. The coupler is also a low loss device,
It typically has an insertion or passage loss of only 2-3 percent. The term "insertion loss" as used herein refers to the loss actually incurred by light passing through the coupler from one end to the other due to scattering. For example, light is given to port A, and 97% of this light is transmitted to port C and (combined) port D.
When reaching , this insertion loss is 0.03 (3%)
becomes. The term "combiner transmission" is defined as 1 minus insertion loss. Therefore, if the insertion loss is 0.03 (3%), the coupler transmission will be 0.97 (97%). Polarization Controller 40 One type of polarization controller suitable for use as polarization controller 40 of FIG.
It is described in Electric Letters Vol. 16, No. 20, pp. 778 to 780, published September 25, 1980. As shown in FIG. 3, the controller includes a base 70 and a plurality of upright blocks 72A-72D mounted thereon.
Spools 74A to 74C are provided between adjacent blocks 72.
They are installed along the tops of shafts 76A to 76C, respectively. This shaft 76 is 1
The blocks 72 are arranged along two axes and are rotatably mounted between blocks 72. Spool 74
is generally cylindrical and positioned along shaft 76, with the axis of spool 74 perpendicular to the axis of shaft 76. fiber part 1
4 (FIG. 1) extend along the bore of the shaft 76 and are wound around each of the spools 74 to form three coils 78A-78C. The radius of the coil 78 is the same as that of the fiber 14.
are compressed to form a birefringent medium in each of the coils 78. This 3
The two coils 78A to 78C are connected to the shaft 74.
A and 74C can be rotated independently of each other about the axes to adjust the birefringence of fiber 14 and thereby control the polarization of light passing through fiber 14. The diameter and number of turns in coil 78 are such that outer coils 78A and 78C provide a quarter-wave spatial delay, while center coil 78B provides a quarter-wave spatial delay. Ordained to give. Quarter wavelength coils 78A and 78C
controls the ellipticity of polarized light, and the half-wavelength coil 7
8B controls the direction of polarization. This allows the polarization state of the waves propagating through the fiber section 14 to be adjusted over a full range. however,
It will be appreciated that this polarization controller may be modified to include only two quarter wave coils 78A and 78C. That is, the direction of polarization (otherwise given by central coil 78B) can be indirectly controlled through appropriate adjustment of the polarization ellipticity using two quarter-wave coils 78A and C. This is because it can also be controlled. Therefore, the polarization controller 40 shown in FIG. 1 includes only two quarter-wave coils 78A and 78C. This arrangement allows the overall size of the controller 40 to be reduced, which may be advantageous when the present invention is applied to specific applications where space is limited. . Thus, polarization controller 40 provides a means for establishing, maintaining, and controlling the polarization state of light propagating through fiber section 14. Operation of the Resonator Referring again to Figure 1, in its operation:
The light entering the fiber portion 12 from the light source 22 is
It propagates to port 1 of coupler 20 where a portion of this light is coupled to port 4 and another portion of the light propagates to port 3. Light at port 4 propagates through fiber section 16 and exits from the end of fiber 10. However, the light at port 3 passes through loop section 14 and enters the coupler again from port 2. This portion is coupled to port 3, while the remaining portion propagates to port 4 and passes through fiber section 16. Loop 14 and coupler 20 cooperate to provide a resonant cavity such that light entering the coupler at port 2 is directed to laser source 2.
It causes interference with the light coming from 2. Such interference causes constructive interference at port 3, while destructive interference at port 4, thereby causing the establishment of light within the resonant cavity loop. Hereinafter, the light from the light source 22 and propagating through the fiber section 12 will be referred to as the input signal wave Wi, while the light exiting the port 4 and propagating through the fiber section 16 will be referred to as the output signal wave Wo. I will call it. The light rotating within the loop portion 14 is called a rotating wave Wc. Since the rotating wave Wc propagates around the loop 14 from port 3 to port 2, part of its power is lost due to transmission loss. “Fiber transmission loss”
The term is defined as the partial loss that occurs while light propagates through the fiber from port 3 to port 2. In the example shown, fiber transmission loss is completely a function of fiber absorption, such that the power or intensity of wave Wc at port 2 is equal to the power of wave Wc at port 3, exp
It is equal to (−2α ₀ L) multiplied by Here, L is the path length of the loop 14 in which the light is rotating,
The phase shift in the coupler 20 is excluded, and α ₀ is the amplitude absorption coefficient of the fiber 10.
If other additional components (eg, fiber optical polarizers) are installed within the fiber loop, the losses due to these components will be included in the definition of fiber transmission loss. The term "fiber transmission" is also defined as the rotating wave power at port 2 divided by the rotating wave power at port 3. Stated another way, this is the amount of power from port 3 that reaches port 2 (ie, fiber transmission = 1 - fiber transmission loss). In addition to being absorbed by fiber transmission losses, this rotating wave Wc is weakened somewhat due to coupler insertion losses in each path through coupler 20. Additionally, the power or intensity of the input wave Wi is lost during propagation through the coupler 20 due to coupler insertion loss. In this regard, coupler 20 can be modeled as a lossless device with summed insertion losses (γ ₀ ) independent of the coupling constant. The relationship between coupler insertion loss and the complex amplitude at each of the four ports of coupling 20 is: |E ₃ | ² + |E ₄ | ² = (1-γ ₀ ) (|E ₁ | ² + | E ₂ | ² ) (1) where; E ₁ , E ₂ , E ₃ and E ₄ are the complex electric field amplitudes at coupler ports 1, 2, 3 and 4, respectively; γ ₀ is the coupler Insertion loss (typically 2%
(on the order of 10%). The complex amplitude at port 3 and port 4 is
The electric field amplitude at port 1 and port 2 has the following relationship: E ₃ = (1-γ ₀ ) ^1/2 (1-K) ^1/2 E ₁ +j√E ₂ (2) and E ₄ = (1 −γ ₀ ) ^1/2 j√E ₁ +(1−K) ^1/2 E ₂ (3) where K is a strong coupling constant. K=0 corresponds to no bond at all, and K=1 corresponds to the presence of a complete bond. Furthermore, there is the following relationship between E ₂ and E ₃ ; E ₂ = E ₃ e ^- 〓 ₀ ^L e ^jBL (4) However; β = nω/C (5) and; α ₀ is the amplitude absorption coefficient of the fiber; L is the length of the fiber loop portion 14; n is the effective refractive index of the fiber; ω is the frequency of the light; β is the propagation constant of the fiber 10; and c is the speed of light. In order to achieve complete resonance, the output wave Wo
must be zero, so the ratio E ₄ /E ₁ must be zero. Therefore, by solving the second, third and fourth equations, E ₄ /E ₁ becomes γ ₀ , K,
Expressed by α ₀ L and βL, and setting E ₄ /E ₁ as 0,
Resonance conditions written in terms of loop length L and coupling constant K are obtained. One of the conditions necessary for resonance is; βL=q2π−π/2 (6) where q is any integer. Therefore, in order to produce perfect resonance,
Loop 14 with phase shift removed by coupler 20
The total phase delay (βL) around must be an integer multiple of 2π radians smaller by π/2. According to the second and third equations, the directional coupler 2
Note that 0 has a phase shift of +π/2. By adding this phase shift to βL in the sixth equation, loop 14 is passed (for example,
It will be seen that the total phase accumulated in the rotating wave Wc when going around the loop from any point in the loop and returning to this arbitrary point is equal to q(2π). As will be appreciated from the discussion below, the length of the loop can be adjusted after the resonator is assembled to satisfy the resonance conditions. This uses an electrically driven PZT cylinder around which the fiber 14 is wound.
This is done by mechanically stretching 4. The resonance conditions specified by Equation 6 can be more fully understood by referring to FIG.
FIG. 4 shows a method of producing constructive interference at port 3 and destructive interference at port 4 by making good use of the π/2 phase shift of coupler 20. For convenience of discussion, combiner 2
0 is shown with an effective coupling point in the central portion of coupler 20 and ports 1, 2, 3, and 4 separated by an integer number of wavelengths from this point. The loop length (L) can be viewed as the distance from this point of attachment, around the loop 14, and back to this point of attachment. This distance is calculated as follows, where q is an integer.
Must be (q-1/4) wavelength. In Figure 4, the input signal wave Wi is a reference wave with phase O, and all other waves (i.e.
It is assumed that the phases of Wc and Wo) are determined relative to this input wave Wi. moreover,
All waves propagating through the coupler 20 are combined into two components, i.e. "cross-coupled", denoted by the subscript c.
component and a “direct” component denoted by the subscript s. Therefore, the input wave Wi is a cross-coupled component Wic that propagates from port 1 to port 4.
and the direct component propagating from port 1 to port 3.
It is divided into Wis and Wis. Similarly, wave Wc is port 2
A cross-coupled component Wcc propagates from to port 3,
Direct component Wcs propagating from port 2 to port 4
It can be divided into If the light source 22 is turned on at t=0, the input wave Wi enters port 1 of the coupler 20 with phase 0 and propagates therein. Cross-coupled component Wic
undergoes a phase shift of +π/2 while propagating to port 4, while the direct component Wis remains unchanged in phase while propagating to port 3.
Therefore, the phase of the light wave Wc at port 3 is 0. This wave Wc then loops 14
It then propagates around and heads towards port 2. If the loop length L is chosen according to equation 6, wave Wc will have a phase of -π/2 when it arrives at port 2. While the wave Wc propagates through the coupler 20, the cross-coupled component Wcc is out of phase by +π/2, and when it arrives at port 3, its phase is 0 as well as the phase of the input wave component Wis. be. Therefore, the rotating wave component Wcc constructively interferes with the input wave component Wis at the port 3, thereby increasing the intensity of the rotating wave Wc at the port 3. On the other hand, the phase of the direct component Wcs of the rotating wave Wc does not change while it propagates from port 2 to port 4, and therefore its phase remains at -π/2 even at port 4. There is. Therefore, this component Wcs destructively interferes with the cross-coupled input wave component Wic having a phase of +π/2. As a result, as wave Wc rotates through loop 14, it is coupled to the input signal wave at port 3.
The power of the wave rotating through the loop 14 ( Strength)
Establish Pc slowly (and asymptotically).
Such a wave is 63% of the equilibrium value (i.e. 1-
The time required to establish up to e) is defined as the cavity growth time (Tc), also commonly referred to as the cavity delay time. To achieve perfect resonance at the equilibrium value and zero output power at port 4,
The second condition must be satisfied. That is, the direct rotating wave component Wcs at port 4 must have an amplitude equal to the amplitude of the cross-coupled input signal component Wic at port 4. In order to generate such a situation, the coupling constant K is adjusted to a value Kr, which will hereinafter be referred to as the "resonant coupling constant". Second
Solving the equation, the third equation, and the fourth equation for E ₄ /E ₁ ,
By setting E ₄ /E ₁ to 0 (this is a resonance condition), this resonance coupling constant Kr becomes; Kr=(1−γ ₀ )exp(−2α ₀ L) ( 7) In the example shown, this combined transmission is (1-γ ₀ ) and the fiber transmission is exp(-2α ₀ L)
It is. Therefore, Kr = coupled transmission x fiber transmission (8). In the disclosed embodiments, the fiber absorption is
8.3dB/Km, loop 14 is 10 meters, and
2α ₀ L equals 0.0158 at a wavelength of 632.8 nm. Since the coupler insertion loss is 1.8%, the resonant coupling constant is 0.967. Using the resonance coupling constant specified by Equation 7, the rotational power (intensity) and output power (intensity) are normalized by the input power by Equations 2, 3, and 4. As the value: |E ₃ /E ₁ | ² = Pc(3)/Pi = (1-γ ₀ ) (
1-Kr) ^{/(1+Kr) 2-4KrSin 2 (} βL/2-π/4)(
9) |E ₄ /E ₁ | ² = Po/Pi = (1-γ ₀ ) [1-(1-Kr)
² / (1 + Kr) ² −4KrSin ² (βL/2−π/4)] (10). Here, Pc(3) is the rotational wave at port 3
is the power (intensity) of Wc; Pi is the input signal wave Wi
is the power (strength) of; and Po is port 4
is the power (intensity) of the output wave Wo at . If βL is chosen to satisfy the resonance condition specified by Equation 6, Equation 9 becomes: |Pc/Pi| _nax =1−γ ₀ /1−Kr (11). This equation can be rewritten as follows. Pi=Pc(1-Kr)+Piγ ₀ (12) If the 6th formula is satisfied, (1-Kr)
is equal to the circulating partial intensity loss of the rotating wave Wc (ie, coupler insertion loss + fiber transmission loss). Therefore, the right-hand side of Equation 12 represents the total power dissipated in coupler 20 and loop 14. As a result, as can be seen from equation 12, in a perfectly resonant condition, the rotational power Pc is such that the total power dissipated in the loop and coupler is equal to the input power Pi at port 1. The normalized theoretical rotational power and output power specified by Equations 9 and 10, respectively, are expressed as 2 in FIGS. 6 and 7, respectively.
It is shown as a function of βL for two typical coupler insertion loss values: 5% and 10%. These curves are for the case where the loop length is 3 meters (2α ₀ L = 0.0057), but similar curves are drawn even when the loop length is 10 meters. You will be able to understand that. As shown in Figure 6, the rotational power Pc strongly depends on the coupler insertion loss, being approximately nine times the input power Pi for a 10% insertion loss, and for a 5% insertion loss. is almost 19 times the input power Pi. In contrast, the output power Po decreases to zero in both cases at full resonance, as shown in FIG. However, the minimum and maximum values in FIGS. 6 and 7 become sharper as the insertion loss decreases, indicating that the cavity finesse is strongly dependent on the coupler insertion loss. This cavity finesse (F) is defined by the following formula. F=FSR/δf (13) where FSR is the free spectral range of the resonant cavity (i.e. the distance between the minimum (Figure 7) or maximum (Figure 6)); and δf is the maximum rotational power. This is the width of the rotational power maximum value (Fig. 6) at one-half of (that is, one-half the power at perfect resonance). This free vector range (FSR) can be defined as follows. FSR=C/nL (14) By setting the ninth equation to be equal to half of |Pc(3)/Pi|max, the full width at half the maximum value becomes: δf= C/nL{1-2/πSin ^-1 [1-(1
−Kr) ² /4Kr〕 ^1/2 }(15) When Kr is close to 1, w can be approximated as follows: The error of this approximation is within 0.2% for Kr greater than 0.8. By substituting equations 14 and 16 into equation 13, the cavity finesse becomes: Recall from equation 8 that the resonant coupling constant (Kr) is equal to the product of coupler transmission and fiber transmission, so (1-Kr) is equal to the total partial loss around loop 14. 17th
It will be seen from the equation that this finesse increases as these partial losses decrease. Therefore, the finesse is strongly loss dependent and will increase as either coupler insertion loss or fiber transmission loss or both are reduced. In the example shown, this finesse is approximately 80, and the free spectral range for a loop 14 length of 10 meters is approximately 20.6 MHz. Finally, referring back to Figure 5, the cavity growth time Tc can be approximated as follows: Tc=nL/C/2(1-Kr) (18) To generate a resonance effect includes a laser light source 20
must have a coherence length greater than c×Tc. Referring to FIG. 8, the resonance effect predicted by equations 9 and 10 shows that the optical power (intensity) of the output wave Wo at the end of the fiber section 16
can be observed by providing a detector 80 that measures . Detector 80 outputs an electrical signal on line 82, which output is proportional to the optical intensity of the output wave Wo. This line 82
is connected to an oscilloscope 84 to input this signal. A signal from a triangle wave generator 86 is provided to this oscilloscope 84 on line 88 and to a phase modulator 90 on line 92. As a particular example, this phase modulator comprises a 3 inch diameter PZT cylinder around which a section of fiber loop 14 is wound only 26 times. The signal from the triangular wave generator 86 is
The PZT cylinder 90 is actuated to spread it radially, so that the fiber 14 stretches linearly to the length (L) of the fiber at the frequency of the generator 86.
will be changed periodically. In this arrangement, the fiber resonator's behavior is somewhat similar to that of a scanning Fabry-Perot interferometer. FIG. 9 is a diagram showing the oscilloscope trajectory of a detector current 96 representing the optical output power (Po) and a triangular wave generator signal 98 representing the amount of stretch of the fiber stretched by the phase modulator 90. be. The amount of fiber stretch provided by signal 98 is slightly greater than the wavelength, so the output power shown in FIG. 9 drops to zero by two degrees between each linear stretch of the fiber. When the coupling constant is changed slightly from the resonant coupling constant Kr, the finesse decreases and a non-zero output power is observed at the minimum of curve 96. The importance of maintaining the polarization state of the light within the fiber loop 14, such as by using a polarization controller 40, is illustrated in FIG.
This figure shows a case where the quarter wavelength loop of the polarization controller 40 is rotated and moved far from the optimum position. As shown therein, two resonant modes corresponding to two independent polarization modes are observed. These two modes resonate at different scan positions because their propagation velocities are slightly different. Each of these resonant modes has a non-zero output power, which means that if one mode is resonant and the other mode is not, then the output power of the non-resonant mode is This is because it is observed at the resonance point of that mode. Operation as an Acoustic Sensor The resonator described above provides a highly sensitive acoustic sensor that detects the frequency and intensity of acoustic waves incident on the fiber loop. When the wavefront of an acoustic wave impinges on the fiber loop 14, the length of this loop is modulated in the same manner as the frequency and intensity of the wavefront. With the length of the loop undergoing such modulation, the full round phase delay through loop 14 is also simultaneously modulated. If the fiber loop cavity is at least partially resonant, modulation in this circular phase delay will cause the output signal to
The optical power of Wo is also modulated in response. Therefore, by detecting the fluctuation in the intensity of the output wave Wo, the intensity and frequency of the acoustic wave can be confirmed. The length of the optical fiber loop is not limited to acoustic waves.
It is also sensitive to other environmental effects, particularly temperature and vibration. However, temperature- and vibration-induced loop changes tend to be at lower frequencies compared to the acoustic band frequencies.
Therefore, the sensor of the present invention detects the low frequency components of the modulated output power and passes these components through a feedback loop to mechanically change the length of the fiber loop 14 to prevent temperature or vibration induced vibrations. This compensates for variations in the length of the loop so that such variations do not affect the detection of the acoustic frequency signal. As shown in FIG. 11, the sensor of the invention includes a detector 110 at the end of the fiber section 16 for detecting the optical power of the output wave Wo. This detector 110 outputs an electrical signal on line 112, which signal is proportional to the optical intensity of the output wave Wo. This line 112 is connected to stabilization electronics 114 to input this signal into this circuit. In response to this signal, stabilization electronics 114 outputs a signal on line 116 that drives a PZT cylinder 118 in fiber loop 114. A signal from stabilization electronics 114 drives PZT cylinder 118 to radially expand it and linearly extend fiber 14 into fiber loop 14.
change the length of In a particular example, this
The PZT cylinder is 3 inches in diameter and has 26 wraps around it at the location of the fiber loop 14. In short, the stabilization electronics 114 provides a bandpass filter that passes low frequency components of the detector current and drives the PZT cylinder 118 to stretch the fiber loop by an amount proportional to such components. Become. This is true for low frequency drift caused by temperature fluctuations, for example.
Stabilization of the total integrated circular phase delay around loop 14 is provided. The stabilization electronics 114 includes a transimpedance amplifier (operational amplifier) 120 for returning the current on line 112 to a voltage. This voltage is output on line 112 to a signal processor 124 which displays or prints the waveform of the sensed acoustic wave, such as from a sound recorder, loudspeaker or paper recorder. It may also be equipped with a display or recording unit. Line 122 is also connected and connected to a low pass filter 126 to provide this signal to this filter.
This filter blocks frequencies greater than approximately 10KHz. The sole purpose of this filter 126 is to prevent the PZT cylinder from resonating, since, as shown in the example, this cylinder has a resonant frequency of about 25 KHz. The output from low pass filter 126 is provided on line 128 to differential amplifier 1
Heading towards 30. This differential amplifier is connected to line 128
is applied to amplifier 13 by line 132.
0 to provide an error signal. The differential amplifier 130 outputs the error signal on line 134 to an integrator 136. This integrator outputs a signal on line 116
Drives the PZT cylinder 118. Through adjustment of this reference voltage, this PZT cylinder 118 is biased such that the loop 14 is at maximum +/-
It operates to be stretched or shortened within a range giving a phase delay of 20 radians. This integrator 136 is an active low pass filter that provides a selective response in gain to low frequencies. This gain is negligible for acoustic frequencies and increases linearly as the frequency of the error signal decreases. Accordingly, stabilization electronics 114 drive PZT cylinder 118 in response to the low frequency drift component of the detector current to account for changes in loop length caused by environmental ambient effects such as temperature and vibration. Compensate. The operation of this acoustic sensor may be more fully understood by analyzing the behavior of the optical output wave Wo, which changes in response to changes in loop length. In this regard, the fiber loop 14
It may help to rewrite equation 10 by the total phase delay around . Here, the phase delay is φ
I will call it. Therefore: Po(φ)=P _nax f(φ) (19) However: P _nax =η(1−γ ₀ ) (20) f(φ)=1−Sin ² (φ/2)/1− ηSin ² (φ/2
)(21) η=4Kr/(1+Kr) ² (22) The term Po used below is the output wave normalized by the intensity of the input wave Wi (i.e. normalized between 0 and 1.0). I would like you to understand this as a demonstration of Wo's strength. Furthermore, the sign Pmax is the output wave
Used to indicate the maximum output power (intensity) of Wo (ie, the optical power at port 4 when the loop is not in resonance). As shown in FIG. 12, Po(φ) fluctuates between Pmax and 0 with a period of 2π. Note that f(φ) varies between 1 and 0 with this same period. The maximum slope of Po (φ) is: |dPo/dφ| _nax = P _nax 1/2 (η−1)Sinφ ₀ /1−ηSin ² (φ ₀ /2)(
23) Here, φ ₀ is the total phase delay corresponding to the point of maximum slope of Po(φ), which is φ ₀ = cos ^-1 {1/η-1/2-1/2√9 −4/η+4/η ² } (24) Referring now to FIG. 12 in conjunction with FIG. 11, the reference voltage on line 132 is preferably
The PZT cylinder 118 is driven to adjust the loop length to provide a phase delay corresponding to φo. This causes the resonator to operate at the point Z of maximum slope on the output power curve 140. The total phase delay φ around the loop 14 at this point Z is equal to an integer multiple of the wavelength plus or minus φo. In the example shown here,
This operating point Z, which corresponds to approximately 0.24 of the normalized output power Po(φ), is determined by the stabilizing electronic circuit 1 described above.
14 stabilizes against low frequency drift. Because the slope of the output power curve 140 is steep,
Due to this small phase shift from the operating point Z,
Substantial variations in output power occur.
Furthermore, because the output power curve 140 is approximately linear around this operating point Z, there is a proportional relationship between small phase shifts and the resulting variations in output power. For example, the output power undergoes a detectable sinusoidal modulation around the operating point Z due to the very small sinusoidal phase shift induced by the change in loop length caused by the acoustic wave. It turns out. For example, as shown in FIG. 13, when the sensed acoustic wave modulates the length of loop 14 such that the integrated phase delay around loop 14 is modulated by Δφ, the output wave Wo
is correspondingly modulated by ΔPo. Although only a sinusoidal modulation of constant amplitude with a single frequency is shown in Figure 13, whatever the intensity and frequency of the acoustic wave, this output wave Wo will be Therefore, it is understood that it varies. Additionally, for illustrative purposes:
Curve 140 is fixed and point Z is on curve 140
Although it is assumed that the curve 140 moves up and down, it should be recognized that in reality, the entire curve 140 shifts around the point R along the X axis while the point Z remains fixed. Such a shift in curve 140 is caused by a change in the length of the loop induced by the acoustic wave, which causes a change in the surface frequency of loop 14. Generally, the sensitivity of this sensor is proportional to the slope of the output power curve. In the embodiment shown, the maximum slope is approximately 14 radian ^-1 , which is a considerable improvement over the maximum slope of 1/2 radian ^-1 of the Mazher-Zender interferometer. The minimum detectable phase shift (i.e., the phase shift when the signal-to-noise ratio is 1) can be defined as: Δφ=ΔTnoise {|dP/dφ| _nax } ⁻¹ where ΔTnoise is the transmission noise (including detection) resulting from the limited noise of the system. In this equation, both Pmax and ΔTnoise are measured in units of photodetector current, not a normalized transmission from 0 to 1. For the value |dP/dφ|max given above, the sensor of the present invention has a theoretical minimum An improvement of approximately a factor 30 in detectable phase shift can be expected. Testing has shown that the main noise source inside the sensor of the present invention is not laser amplitude noise or detector shot and temperature noise, but rather
It turned out to be frequency noise from the laser light source 22. For example, as long as the laser 22 is not strictly monochromatic and varies from its frequency at the operating point Z,
This laser can generate frequency noise that encompasses the acoustic frequency band. Therefore, this frequency noise can be detected and misinterpreted as acoustic wave induced variations over the length of the fiber loop. For a given loop length and free spectral range, this frequency noise, when detected in the optical output wave Wo, will correspond to the slope (dP/dφ) of the output power curve 140 for a given loop length. It has a proportional size. Therefore, for this particular type of noise, the signal-to-noise ratio does not depend on the slope of curve 140. It is believed that such noise can be reduced by improving the monochromaticity of laser light source 22. The signal-to-noise ratio caused by laser frequency noise is independent of dP/dφ, but this noise ratio can be reduced by shortening the loop length and increasing the free vector range of the resonator. be able to. Increasing the free spectral range in this way has a small effect on dP/dφ, but when f is the frequency, it produces an effective effect on dP/df.
For example, when the finesse of the resonator is 80,
Changing the phase by 0.1π corresponds to a frequency change of approximately 1 MHz for a free spectral range of approximately 20 MHz, and a frequency change of 0.2 MHz for a free spectral range of 4 MHz. Become. Therefore, a loop with a free spectral range of 20 MHz will be less sensitive to laser frequency noise than a loop with a free spectral range of 4 MHz, given the same laser frequency noise. Become. As a result, a highly monochromatic light source with a long coherence length is desirable in applications where long loop lengths are important. However, to reduce noise problems, it is better to keep the loop length as short as possible. In practice, the loop length is such that the amount of optical fiber exposed to the sensed quantity is sufficient to ensure that the optical output signal is measurable within the desired dynamic range. selected. In the example shown here, the maximum dynamic range is from 0 to
It will be 0.48 (normalized power: Figure 13).
This is because the operating point Z is 0.24 (normalized power) on the curve 140. In the graph of FIG. 14, the overall response of the stabilized interferometric sensor as a function of frequency is shown. The average output power measured by detector 110 is normalized between 0 and 1 to be Y
The sensed acoustic wave frequencies are plotted on the X-axis to form a response tune 144. Below 100Hz, this output power is essentially absorbed. This is because, in this frequency range, loop fluctuations caused by sensed acoustic wave waves are indistinguishable from those due to temperature drift. However, above 1KHz this absorption does not occur. Therefore, the sensor of this invention
It has a flat response for frequencies above 1KHz. However, in an environment where temperature drift is not a problem, such a curve 14
4 is not absorbed by the stabilization system, so that even infrasound waves can be sensed.

[Brief explanation of the drawing]

FIG. 1 is a schematic diagram of a resonator using the Brillouin ring laser of the present invention, showing a light source for introducing light into the fiber loop and a fiber optical directional coupler for closing the fiber loop. There is. FIG. 2 is a cross-sectional view of one embodiment of a fiber optical directional coupler used in the resonator of FIG. Figure 3 shows one of the fiber optic polarization controllers used in the resonator of Figure 1.
FIG. 3 is a perspective view of two embodiments; FIG. 4 is a schematic diagram of the fiber optic directional coupler shown in FIGS. 1 and 2, showing the phases of light wave components propagating therein. FIG. 5 is a graph showing rotational power as a function of time, showing the rotational power growing asymptotically to an equilibrium value after a time interval equal to the cavity growth time. Figure 6 shows port 3 as a function of phase delay through the fiber loop, normalized by input power.
2 is a graph of the rotational power at the resonance point, showing the normalized rotational power at the resonance point when the coupler insertion loss is 5% and 10% as an example. Figure 7 is normalized by input power.
Graph of output power as a function of phase lag through the fiber loop, showing zero output power at resonance for example coupler insertion losses of 5% and 10% .
8 is a schematic diagram of one embodiment of the resonator of FIG. 1; FIG. FIG. 9 is a graph showing the resonance behavior for the embodiment shown in FIG. FIG. 10 is a diagram similar to FIG. 9, and is a graph showing the effect of light propagation in both resonance modes when the polarization controllers are arranged incorrectly in resonance behavior. FIG. 11 is a schematic diagram of the acoustic sensor of the present invention using the resonator of FIG. The stabilization feedback electronics that provide the stabilization are shown. FIG. 12 is a graph of the optical output power at port 4 as a function of the round-trip phase delay of the light rotating through the fiber loop, showing the point of maximum slope at which the sensor operates in a stabilized manner. . FIG. 13 is a diagram showing a portion of the graph of FIG. 12, illustrating the effect on optical output power of phase modulation caused by acoustic wave-induced loop length variations; ing. FIG. 14 is a graph plotting the average optical output signal power (normalized between 0 and 1.0) as a function of the frequency of the sensed acoustic wave, illustrating the response of the sensor of the present invention. It shows. 132 represents a stabilizing electronic circuit.

Claims (1)

Claims: 1. A single seamless strand of optical fiber having first and second ends and a loop portion therebetween; a fiber optical coupler optically closing said fiber loop; a light source providing a light wave Wi at the first end, the light wave propagating through the fiber to the second end, and an optical output having an intensity dependent on the length of the loop. further comprising a phase delay device for providing a signal Wo to control a circular phase delay of light rotating through the fiber loop and providing a resonant cavity for the rotating light; sensing a second environmental effect, the first environmental effect varying in a first frequency range, and the second environmental effect varying in a second frequency range; driving the phase delay device; further comprising a stabilizing device to compensate for changes in fiber loop length caused by the first environmental effect to stabilize the output signal Wo with respect to the first environmental effect, the output signal Wo being The fiber optical sensor remains sensitive to the second environmental effect and further comprises: a detector for detecting a change in the intensity of the output signal Wo to detect the second environmental effect. 2. The phase delay device provides the rotating light with the circular phase delay equal to an integer multiple of the wavelength ±φ ₀ , where: φ ₀ =cos ⁻¹ {1/η−1/2−1/2 √9-4/
The fiber optical sensor of claim 1, wherein η+4η ² } (25) and η=4Kr/(1+Kr) ² (26) where Kr is a resonant coupling constant. 3. The phase delay device 118 provides a full-circle phase delay to the rotating light so that the sensitivity of the output signal Wo to changes in loop length is maximized. Fiber optical sensor as described in Section. 4. A fiber optical sensor according to any one of claims 1 to 3, wherein the first environmental effect is temperature. 5. A fiber optical sensor according to any one of claims 1 to 4, wherein the second environmental effect includes an acoustic wave. 6. The phase delay device is wound around a portion of the fiber loop.
A fiber optical sensor according to any one of claims 1 to 5, comprising a PZT cylinder. 7. The stabilization device: comprises an electronic circuit for driving the phase delay device with an electrical signal from the detector, the electronic circuit filtering the signal to be within the first frequency range. 7. A fiber optic sensor as claimed in any one of claims 1 to 6, which passes frequency components and effectively blocks frequency components lying within said second frequency range. 8. The fiber optical sensor of claim 7, wherein said electronic circuitry comprises an integrator that provides selective gain for signal components within said first frequency range. 9. A method for detecting acoustic waves using an optical fiber, comprising the steps of forming a loop in the optical fiber; propagating input light through said loop to provide an optical output signal Wo; fiber optical coupling. closing the loop using a resonator; and selecting a length of the loop to provide a resonant cavity, the loop length being determined by the output signal.
affecting the strength of the output signal Wo, further comprising the step of compensating for thermally induced loop length changes to stabilize the strength of the output signal Wo with respect to the thermally induced changes. and detecting the presence of the acoustic waves by detecting the output signal Wo. 10 The steps for making the selection include: determining a maximum point of sensitivity of the intensity of the output signal Wo to changes in loop length; and selecting the loop length to provide an output signal strength at the maximum sensitivity point. prepare,
A method for sensing acoustic waves using the optical fiber according to claim 9. 11 The compensation step includes: detecting a component of the output signal Wo having a frequency lower than the frequency of the sensed acoustic wave; and changing the length of the loop in response to the detected frequency component. A method for sensing acoustic waves using an optical fiber according to claim 9 or claim 10, comprising the steps of: